Pregnant Sprague-Dawley CD rats were orally administered either phenytoin (PHT, 200 mg/kg), mephenytoin (MPH, 100 mg/kg), ethotoin (ETH, 600 mg/kg), hydantoin (HYD, 1,200 mg/kg) or vehicle (propylene glycol) on days 7-18 of gestation. Mean (+/- S.E.) maternal serum concentrations of PHT, MPH, and ETH 1 hour after dosing on gestational day 18 were 16.0 +/- 3.3, 10.7 +/- 3.0, and 65.2 +/- 10.45, respectively, and free fractions were 16%, 18%, and 11% respectively. The free fraction for PHT is similar, but was lower for both MPH and ETH than that seen in humans. Preweaning mortality for PHT, MPH, ETH, HYD, and controls was 25%, 6.3%, 12.5%, 2.0% and 0.8%, respectively. The MPH and ETH-exposed animals weighed approximately 6.6% less than controls throughout the study; the other groups did not differ significantly. PHT offspring showed increased early locomotor activity. Only PHT-exposed animals (27%) exhibited abnormal circling behavior after weaning. PHT-circlers accounted for higher levels of activity in an open-field test and for longer straight channel swimming times. PHT-circlers and noncirclers differed from one another and controls on performance of a complex (Cincinnati) maze and on the development of the air-righting reflex. Offspring prenatally exposed to MPH showed an early delay in air-righting. ETH and HYD offspring were not consistently different from controls in behavior. The data suggest the following ordinal relationship among the drugs for behavioral teratogenesis: PHT much greater than MPH greater than ETH congruent to HYD congruent to CON. The effects of PHT are consistent with previous findings. Data on the other drugs suggest that other hydantoins do not possess the behavioral teratogenic efficacy of PHT and that PHT may be unique in its effects on CNS development.